Salt Peach

Marketing mix modeling is a technique that takes spend data along with other inputs to predict future sales. It is a powerful tool in the marketers tools kit because as every marketer knows, marketing strategy is complex. You are dealing with multiple channels from more traditional sources such as TV or print to a vast array of digital channels like display, social and paid search. Each channel as a role, and no channel operates in isolation. The power behind marketing mix modeling is that you can bring all of your different channel spends under one umbrella to make a more accurate prediction of future sales. Because you are able to build a model that uses multiple different channel spends you get a more robust understanding of what you can expect as a result of your marketing efforts.

To put it simply, marketing is complex. There are many aspects of marketing that are correlated, although proving causation may not be possible. Take for example advertising on the Google Display Network. This is typically used as an awareness tactic. Display helps keep your product top of mind, but many businesses will not see a large number of conversions from this channel. On the other hand it is also cheap, usually costing pennies on the dollar when compared with a high intent channel like paid search. Let’s not forget that many potential customers will not even click on your display ad, but that does not mean that there was no impact. How do we quantify the new customer who google searched our brand when they got home because they saw a display ad at work that kept our brand top of mind? If we are talking about direct attribution, we can’t. There is no digital connection between the two events. However there is a way to better understand the complex impact of multichannel spend tactics, and that is by using marketing mix modeling.